96
H. K. Palo
Data Model provide semantic and syntactic information. Nevertheless, it cannot find
a general framework for any large-scale. Ontology can be viewed as the formal
specification of any concept. It provides the working model on the properties, relationships, types, and interactions among different entities in an IoT system. It helps
the SIoT by facilitating the context or reference on a wide range of vocabularies.
Its major components are (a) class (b) individuals (c) attributes (d) relations [24]
and may not be limited to physical objects only. A few popular ontologies are
SSN (Semantic Sensor Network) Ontology, IoT-O Ontology, Thing Description,
etc. Semantic interoperability allows the exchange of unambiguous and machineunderstandable data. The use of semantic level facilitates intelligent data federation,
data inference, and true machine intelligence. However, it is difficult to provide the
desired semantic understanding without modeling them uniformly with ontologies.
Ontologies utilize the vocabularies as well as the taxonomies or the terminologies to
represent the properties, concepts, and interrelations among data meaning. As ontologies are machine-understandable these are crucial for the inter-machine automation
to provide a high value to the IoT domain.
Ontology can capture reusable information in a domain and is a specification
of a conceptualization. Web Ontology Language (OWL) is an ontology language
widely applied in the field of Semantic IoT. With the growing demand for efficient SIoT, much large-scale ontology is developed in real-time to occupy and integrate information and knowledge. It is essential to focus on the complexity of these
ontologies for clear understanding, reusability, and integration. Based on the software metric concept, a set of ontologies at the class-level and ontology-level have
been investigated to study the design complexities [10]. The ontology metrics are
evaluated against Weyuker’s criteria with empirical analysis of the characteristics and
applicability. Thus, SW ontologies can facilitate the integration and management of
knowledge for autonomous processing of the data by software professionals. There
have been many ontology languages such as OWL, RDF, RDF Schema, etc. that
provide the much-needed vocabularies to represent the essential domain knowledge,
data aggregation and integration from different sources. These languages can only
provide the conceptual data rules and models, but not the specific serialization format.
Some other specific languages such as Turtle, N3 (Notation3), JSON-LD (Javascript
Object Notation for Linked Data), extensible Markup Language (XML), N-Triples
can describe semantic data [36]. The SIoT has a standard query language known as
SPARQL.
The ontologies in SIoT provide a common language to describe objects or things
and their interrelationships. In this regard, the SSN Ontology designed by the W3C
SSN Incubator Group helps to represent the properties of sensors for desired observations and can be extended to generate new features in SIoT. The NCI Thesaurus
Ontology is an OWL ontology that applies SW to develop mutually agreeable
and consistent vocabularies to contain and integrate domain knowledge from many
sources. It has nearly 60,000+ named classes, an approximately similar number of
anonymous classes, and 100,000+ properties (connections) among these classes. This
way, the ontology can represent information on approximately 8000 therapies and
10,000 cancers. Similarly, using the Linked Data project, it is possible to generate and
H. K. Palo
Data Model provide semantic and syntactic information. Nevertheless, it cannot find
a general framework for any large-scale. Ontology can be viewed as the formal
specification of any concept. It provides the working model on the properties, relationships, types, and interactions among different entities in an IoT system. It helps
the SIoT by facilitating the context or reference on a wide range of vocabularies.
Its major components are (a) class (b) individuals (c) attributes (d) relations [24]
and may not be limited to physical objects only. A few popular ontologies are
SSN (Semantic Sensor Network) Ontology, IoT-O Ontology, Thing Description,
etc. Semantic interoperability allows the exchange of unambiguous and machineunderstandable data. The use of semantic level facilitates intelligent data federation,
data inference, and true machine intelligence. However, it is difficult to provide the
desired semantic understanding without modeling them uniformly with ontologies.
Ontologies utilize the vocabularies as well as the taxonomies or the terminologies to
represent the properties, concepts, and interrelations among data meaning. As ontologies are machine-understandable these are crucial for the inter-machine automation
to provide a high value to the IoT domain.
Ontology can capture reusable information in a domain and is a specification
of a conceptualization. Web Ontology Language (OWL) is an ontology language
widely applied in the field of Semantic IoT. With the growing demand for efficient SIoT, much large-scale ontology is developed in real-time to occupy and integrate information and knowledge. It is essential to focus on the complexity of these
ontologies for clear understanding, reusability, and integration. Based on the software metric concept, a set of ontologies at the class-level and ontology-level have
been investigated to study the design complexities [10]. The ontology metrics are
evaluated against Weyuker’s criteria with empirical analysis of the characteristics and
applicability. Thus, SW ontologies can facilitate the integration and management of
knowledge for autonomous processing of the data by software professionals. There
have been many ontology languages such as OWL, RDF, RDF Schema, etc. that
provide the much-needed vocabularies to represent the essential domain knowledge,
data aggregation and integration from different sources. These languages can only
provide the conceptual data rules and models, but not the specific serialization format.
Some other specific languages such as Turtle, N3 (Notation3), JSON-LD (Javascript
Object Notation for Linked Data), extensible Markup Language (XML), N-Triples
can describe semantic data [36]. The SIoT has a standard query language known as
SPARQL.
The ontologies in SIoT provide a common language to describe objects or things
and their interrelationships. In this regard, the SSN Ontology designed by the W3C
SSN Incubator Group helps to represent the properties of sensors for desired observations and can be extended to generate new features in SIoT. The NCI Thesaurus
Ontology is an OWL ontology that applies SW to develop mutually agreeable
and consistent vocabularies to contain and integrate domain knowledge from many
sources. It has nearly 60,000+ named classes, an approximately similar number of
anonymous classes, and 100,000+ properties (connections) among these classes. This
way, the ontology can represent information on approximately 8000 therapies and
10,000 cancers. Similarly, using the Linked Data project, it is possible to generate and